The stress of caring for a loved one with chronic illness has been associated with childhood obesity. Hair cortisol has been proposed as a novel biomarker of chronic psychological stress.
Chen et al BMC Pediatrics (2015) 15:9 DOI 10.1186/s12887-015-0322-y RESEARCH ARTICLE Open Access Caregivers’ hair cortisol: a possible biomarker of chronic stress is associated with obesity measures among children with disabilities Xiaoli Chen1*, Bizu Gelaye1, Juan Carlos Velez2, Clarita Barbosa2, Micah Pepper2, Asterio Andrade2, Wei Gao3, Clemens Kirschbaum3 and Michelle A Williams1 Abstract Background: The stress of caring for a loved one with chronic illness has been associated with childhood obesity Hair cortisol has been proposed as a novel biomarker of chronic psychological stress This study aimed to evaluate the associations between caregivers’ chronic stress evaluated by hair cortisol concentrations (HCC) and obesity measures among children with disabilities such as autism Methods: Eighty-five dyads of children with disabilities and their primary caregivers participated in the study between April and July 2013 in the Patagonia Region, Chile Trained research staff conducted anthropometric measurements of children and caregivers Cortisol concentrations, extracted from hair samples with methanol, were quantified using liquid chromatography tandem mass spectrometry Pearson’s correlation coefficients and linear regression models were used to examine the associations between caregiver HCC (log-transformed) and child obesity measures with adjustment for covariates Results: Caregiver HCC were positively and significantly correlated with child weight (child age- and sex-adjusted r =0.23, P = 0.036), body mass index (BMI) (r = 0.23, P = 0.035), circumferences of neck (r = 0.30, P = 0.006), waist (r = 0.27, P = 0.014), and hip (r = 0.22, P = 0.044) After adjustment for children’s age and sex, caregiver HCC were significantly related to child weight (kg) (beta = 4.47, standard error (SE) = 2.09), BMI (kg/m2) (beta = 1.52, SE = 0.71), neck circumference (cm) (beta = 1.20, SE = 0.43), waist circumference (cm) (beta = 3.75, SE = 1.50), and hip circumference (cm) (beta = 3.02, SE = 1.48) Caregiver HCC were also positively but not statistically significantly associated with child waist-to-hip ratio (beta = 0.01, SE = 0.01; P = 0.191) or body fat percentage (%) (beta = 2.11, SE = 1.28; P = 0.104) Further adjustment for other covariates including child disability diagnosis and caregiver age, sex, education, current smoking, perceived stress, and caregiver BMI did not change the results substantially Conclusions: Chronic stress of caregivers, evaluated by increased cortisol concentrations in hair, was positively associated with obesity measures among children with disabilities Keywords: Hair cortisol, Chronic stress, Adiposity, Child, Disability, Caregiver * Correspondence: xchen@hsph.harvard.edu Department of Epidemiology, Harvard School of Public Health, Boston, MA, USA Full list of author information is available at the end of the article © 2015 Chen et al.; licensee BioMed Central This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated Chen et al BMC Pediatrics (2015) 15:9 Background Children with disabilities have higher prevalence of obesity than children without disabilities [1-7] The 2008-2010 National Health Interview Survey showed that the prevalence of obesity among US adolescents aged 12-17 years with developmental disabilities was 20.4% as compared with 13.1% of adolescents without developmental disabilities Among adolescents with developmental disabilities, those with autism had the highest prevalence of obesity (31.8%) [6] Apart from unhealthy lifestyle factors (e.g., physical inactivity, poor diet), the influence of chronic stress on childhood obesity has been increasingly recognized [8-12] Several studies have indicated a positive association between parental stress and child obesity [9,10,13] Caregivers of children with disabilities represent a population that is known to have high levels of chronic stress [14-16] In addition to traditional parenting responsibilities, caregivers of disabled children must also fulfill disabilityrelated caregiving needs and therefore are prone to chronic stress [14,17] Identifying chronic stress levels among caregivers of children with disabilities may be critical in developing effective intervention and prevention strategies to reduce childhood obesity Stress scales or questionnaires have been widely and typically used to capture short-term subjective stress levels, which can contribute to reporting errors [10,18] Hair cortisol has been proposed as a novel biomarker of chronic stress that has been recently recognized as the most promising way to measure long-term cortisol synthesis and secretion for periods of several months [19-22] Hair collection is simple and non-invasive for participants, and hair cortisol concentrations (HCC) are not influenced by moment-to-moment variations compared with other measures of cortisol from blood, saliva, or urine [20] Body mass index (BMI) is the most commonly used measure of adiposity to describe general obesity Because BMI does not adequately describe regional or central adiposity, other indices of body fatness such as neck circumference and waist circumference have being explored to evaluate central obesity Neck circumference is an emerging measure of central obesity and obstructive sleep apnea [23-26] It has been reported that long-term HCC are increased in shift workers and positively associated with BMI [27] To date, no research has examined whether caregivers’ chronic stress evaluated by hair cortisol is associated with obesity measures among children with disabilities This study aimed to fill the research gap by examining the associations between caregiver HCC and disabled children’s obesity measures including weight, BMI, circumferences of neck, waist, and hip, waist-to-hip ratio (WHR), and body fat percentage We hypothesize that caregiver HCC are positively associated with child obesity measures Page of 13 Methods Participants Between April and July 2013, the Chile Pediatric and Adult Sleep and Stress Study (CPASS) was conducted with the inclusion of hair sample collection among children with disabilities and their caregivers at the Centro de Rehabilitacion Club de Leones Cruz del Sur in the Patagonia Region, Chile Details about the study design have been described elsewhere [28] Briefly, using a recruitment script, research staff approached primary adult caregivers when they checked in for their children’s appointment at the center A total of 129 caregivers of children with physical and/or mental disabilities (e.g., cerebral palsy, autism) were invited to participate in the study Ninety caregivers (including caregivers with each having eligible children) with 93 children (72%) agreed to participate and enrolled in the current study Interviewer-administered questionnaires were used to collect information from primary caregivers about children’s and caregivers’ sociodemographic and lifestyle factors, as well as caregivers’ perceived stress Children’s electronic medical records were reviewed for the confirmation of disability diagnoses and medication use By following standardized procedures, trained research staff took anthropometric measurements twice and collected hair samples from both children and caregivers Among enrolled participants, hair samples from 97.8% of children (two children had shaved head without hair samples collected) and from 98.9% of caregivers (one caregiver had shaved head without hair sample collected) were collected Hair samples from children and caregivers were excluded because of insufficient amount of specimen Hence, a total of 87 children and 86 caregivers (95.6% of enrolled child-caregiver dyads) completed the study protocol and were included in the present analysis Anthropometric measurements Weight (kg) and height (cm) were measured when children and caregivers were wearing light clothing without shoes Height was measured with a telescopic height measuring instrument (Seca 225, Seca Ltd) to the nearest 0.1 cm Weight and body fat percentage were measured with a bioelectric impedance analysis (BIA) scale (Tanita® BC 420 SMA; Tanita Europe GmbH) Weight was measured to the nearest 0.1 kg, while body fat percentage was measured to the nearest 0.1% Circumferences (cm) of neck, waist, and hip were measured using an inelastic tape (Seca 200, Seca Ltd) to the nearest 0.1 cm, with participants in a standing position The averages of anthropometric measurements were calculated and used in the data analysis BMI was calculated by dividing weight (kg) by height squared (m2) WHR Chen et al BMC Pediatrics (2015) 15:9 was calculated as the ratio of waist circumference divided by hip circumference Based on the Centers for Disease Control and Prevention (CDC) growth charts [29], children’s age- and sex-specific BMI was calculated and defined children’s overweight (85th ≤ BMI < 95th percentile) and obesity (BMI ≥ 95th percentile) For caregivers, the World Health Organization (WHO) criteria were used to define normal weight (BMI < 25 kg/m2), overweight (25 ≤ BMI < 30 kg/m2), and obesity (BMI ≥ 30 kg/m2) [30] Hair sample collection and hair analysis A detailed description of the methods used to measure HCC can be found elsewhere [22,31] In brief, hair samples were cut from the posterior vertex of the scalp, as close to the scalp as possible The most proximal cm of the hair strands were used, corresponding roughly to a period of months Hair cortisol extraction procedures were similar to the methods detailed by Stalder et al [32] with some modifications as below Hair samples were washed in 2.5 mL isopropanol for minutes and dried for at least 12 hours, after which 7.5 ± 0.5 mg of whole, non-pulverized hair was weighed out Centrifugation was omitted since whole hair was used Hair samples were incubated in 1.8 mL methanol for 18 hours at room temperature, and then 1.6 mL of clear supernatant was transferred into a glass vial Subsequently, methanol was evaporated at 55°C under a steady stream of nitrogen The residue was re-suspended using 150 μl distilled water + 20 μl of internal standard (cortisol-d4), 150 μl of which was used for liquid chromatography tandem mass spectrometry (LC-MS/MS) analysis Intra-assay and inter-assay coefficients of variance were between 3.7% and 9.1% [31] Covariates Child-specific factors included sex, age (years), disability diagnosis, medication use, and caregiver-reported lifestyle behaviors including diet quality, caffeinated beverage consumption, screen time, and sleep duration According to the International Classification of Diseases (ICD-10) [33], the diagnoses of children’s disabilities were categorized as the following groups: 1) Mental and behavioral disorders, such as autism, attention deficit hyperactivity disorder, and mental retardation; 2) Diseases of the musculoskeletal system and connective tissue, skin and subcutaneous tissue, such as scoliosis; 3) Diseases of the nervous system, such as cerebral palsy; and 4) Congenital malformations, deformations and chromosomal abnormalities, such as Down syndrome Of note, in this study, no children or caregivers had Cushing’s syndrome, a disease characterized by hypercortisolism Page of 13 Caregiver-specific factors included sex, age (years), caregiver-child relationship, marital status, education level, current smoking status, hair-related traits, perceived stress, and obesity measures Caregivers reported their hair color, hair structure (straight or curly hair), and the use of hair treatment including coloration, bleaching, and permanent wave Perceived stress was measured using the 14-item Perceived Stress Scale (PSS-14), which includes ratings of feeling overwhelmed, out of control, and stressed over the past month The PSS-14 has been validated and used widely [34-36] The PSS-14 total score ranges from to 56, with higher scores indicating higher levels of perceived stress In this study, the Cronbach alpha coefficient of the PSS-14 was 0.76, indicating that the PSS-14 had good internal consistency [37] Caregivers in the upper quartile of the PSS-14 score ≥ 27 were considered as having higher perceived stress, while those in other quartiles (PSS-14 score < 27) were considered as having lower perceived stress Statistical analysis As this study focused on caregiver HCC and child obesity measures, we included 86 dyads of children and caregivers with complete hair data and obesity measures One parent with extreme outlying HCC exceeding three interquartile ranges from the median and was excluded from the data analyses, leaving a final analyzed sample of 85 dyads of children and caregivers As some children diagnosed with cerebral palsy and other disabilities were unable to stand on the BIA scale, only 70 children had measured data for body fat percentage Note that for three families with two eligible children at each family, we chose to include one child from each family who were first enrolled in this study We also conducted sensitivity analysis by excluding participants from these families and found similar results (data not shown) We first conducted Kolmogorov-Smirnov tests to determine the normality of caregiver HCC (exposure variable) and child obesity measures (outcome variables) As caregiver HCC were not normally distributed (Kolmogorov-Smirnov test: P < 0.05), HCC were logarithmically transformed to attain normal distribution and used in the data analyses For descriptive purposes, we provided information on means in original units of HCC (pg/mg) Child and caregiver characteristics were presented as means and standard deviations (SDs) for continuous variables and percentages for categorical variables Student’s t-tests were conducted to evaluate the differences in continuous variables including age and obesity measures (e.g., BMI) by sex for both children and caregivers Chisquare tests or Fisher’s exact tests were conducted to evaluate the differences in categorical variables including disability diagnosis, medication use, sociodemographic Chen et al BMC Pediatrics (2015) 15:9 and lifestyle factors, hair traits, and perceived stress by sex among children and caregivers Analysis of variance or Student’s t-tests were used to assess the differences in caregivers’ log-transformed HCC across child disability diagnosis, medication use, sociodemographic and lifestyle factors, caregiver hair-related traits, perceived stress, and obesity status Pearson’s correlation coefficients were calculated to examine the correlations between caregiver log-transformed HCC and child weight, BMI, neck circumference, hip circumference, waist circumference, WHR, and body fat percentage with and without adjustment for child age and sex Linear regression models were fitted to examine the associations between caregiver HCC and child obesity measures, with and without adjustment for covariates from both children and caregivers For each outcome variable, model was unadjusted; model was adjusted for child age and sex; model was further adjusted for child disability diagnosis and caregiver age, sex, education level, current smoking status, and perceived stress Additional adjustment for caregiver BMI and hairrelated factors including hair color and hair treatment did not change the results substantially (data not shown) By calculating Pearson’s correlation coefficients of caregiver perceived stress with caregiver HCC and child obesity measures, we conducted exploratory data analysis to determine whether caregiver perceived stress evaluated by the PSS-14 was correlated with caregiver HCC and child obesity measures The significance levels were set at alpha < 05 and all reported P values are two-sided All statistical analyses were performed using SAS® version 9.3 (SAS Institute, Inc, Cary, NC) Protection of study participants This study was approved by the institutional review boards of Centro de Rehabilitacion Club de Leones Cruz del Sur in Punta Arenas, Chile and Harvard School of Public Health, USA Because children with developmental delays such as mental retardation and motor/speech delays that would affect their ability of providing informed consent, only parents/legal guardians provided the consent for this study Results Characteristics of study participants Table shows the descriptive characteristics of children and caregivers The mean age of children with disabilities (boys: 42.4%) was 15.4 (SD: 2.8) years There were no significant differences in age, disability diagnosis, medication use, lifestyle factors, weight, neck circumference, or waist circumference between boys and girls Boys had significantly higher WHR than girls, whereas girls had higher BMI, hip circumference, and body fat Page of 13 percentage than boys Based on the CDC 2000 criteria, 23.5% of children were overweight and 22.4% were obese Distributions of caregiver HCC across child and caregiver characteristics As shown in Table 2, there were no statistically significant differences in caregiver HCC across child sex, age, disability diagnosis, medication use, or lifestyle factors Caregiverchild relationship, caregivers’ sociodemographic factors, smoking status, hair traits including hair color and hair treatment, perceived stress, or obesity status were not significantly related to HCC among caregivers (all P > 0.05) In total, 55.3% of caregivers were overweight while 35.3% were obese There was no statistically significant difference in HCC among caregivers with normal weight, overweight, and obesity, although caregivers with overweight and obesity tended to have higher levels of HCC as compared to caregivers with normal weight Associations between caregiver HCC and child obesity measures Figure shows the scatter plots of caregivers’ HCC by children’s obesity measures There were significant correlations between caregiver HCC and child weight (Figure 1a) (r = 0.25, P = 0.023), BMI (Figure 1b) (r = 0.23, P = 0.037), neck circumference (Figure 1c) (r = 0.32, P = 0.003), waist circumference (Figure 1d) (r = 0.28, P = 0.009), and hip circumference (Figure 1e) (r = 0.23, P = 0.037) Caregiver HCC were positively but not significantly related to child WHR (Figure 1f) (r = 0.14, P = 0.193) or body fat percentage (Figure 1g) (r = 0.16, P = 0.188) As shown in Table 3, after adjustment for child age and sex, caregiver HCC were positively and significantly related to child weight (r = 0.23, P = 0.036), BMI (r = 0.23, P = 0.035), circumferences of neck (r = 0.30, P = 0.006), waist (r = 0.27, P = 0.014), and hip (r = 0.22, P = 0.044) However, caregiver HCC were positively but not statistically significantly associated with WHR (r = 0.15, P = 0.191) or body fat percentage (r = 0.20, P = 0.104) Linear regression models show that caregiver HCC were positively associated with child obesity measures including weight, BMI, and circumferences of neck, waist, and hip (Table 4) Caregivers’ HCC were also positively but not significantly related to WHR or body fat percentage After adjustment for children’s age and sex, caregiver HCC were positively and significantly related to child weight (kg) (beta = 4.47; standard error (SE) = 2.09), BMI (kg/m2) (beta = 1.52; SE = 0.71), neck circumference (cm) (beta = 1.20; SE = 0.43), waist circumference (cm) (beta = 3.75; SE = 1.50), and hip circumference (cm) (beta = 3.02, SE = 1.48) These associations persisted after further adjustment for child disability diagnosis and caregiver age, sex, education, current smoking, and perceived stress Chen et al BMC Pediatrics (2015) 15:9 Page of 13 Table Characteristics of 85 dyads of children with disabilities and their caregivers Characteristics Total Males Females Mean (SD)/% Mean (SD)/% Mean (SD)/% P value Child characteristics n = 85 n = 36 n = 49 Age (years), mean (SD) 15.4 (2.8) 15.0 (2.4) 15.6 (3.1) 0.339 Mental and behavioral disorders 29.4 30.6 28.6 0.820 Diseases of musculoskeletal system & connective/skin tissue 23.5 27.8 20.4 Diseases of the nervous system 30.6 27.8 32.6 Congenital malformations & chromosomal abnormalities 16.5 13.9 18.4 Disability diagnosis1, % Medication use, % 41.2 50.0 34.7 0.157 Poor diet quality, % 30.6 38.9 24.5 0.155 Caffeinated beverage consumption, % 57.7 61.1 55.1 0.580 Screen time ≥ hours/day, % 54.8 60.0 51.0 0.506 Sleep duration < hours, % 32.9 38.9 28.6 0.600 Weight (kg), mean (SD) 52.9 (15.3) 52.1 (17.8) 53.4 (13.3) 0.684 BMI (kg/m ), mean (SD) 22.8 (5.2) 21.4 (4.4) 23.9 (5.6) 0.029 Neck circumference (cm), mean (SD) 34.3 (3.5) 35.1 (3.7) 33.7 (3.2) 0.070 Waist circumference (cm), mean (SD) 78.2 (11.0) 78.0 (11.0) 78.4 (11.2) 0.874 Hip circumference (cm), mean (SD) 91.5 (11.3) 88.4 (11.1) 93.7 (11.1) 0.031 Waist-to-hip ratio, mean (SD) 0.86 (0.07) 0.88 (0.07) 0.84 (0.07) 0.003 Body fat percentage (%), mean, SD 27.4 (9.6) 22.5 (9.4) 30.7 (8.5)